Why legal is the hardest sector to measure and the most valuable to get right
Legal search is the most expensive advertising real estate that exists. Clicks in personal injury, in some practice areas, cost more than a good meal. A modest paid budget disappears in days, and the client knows precisely what they are spending because it is one of their largest line items.
What makes it hard is not the cost — it is the variance underneath it. A single catastrophic injury case can be worth more than a hundred straightforward matters. Twenty enquiries producing one significant case can be a spectacular month; two hundred enquiries producing none is a disaster. Cost per lead, which works reasonably well in sectors where transactions are broadly similar, is close to meaningless here. It can move in the opposite direction to the firm's actual revenue for months at a time.
That mismatch is where agency relationships in this sector go wrong. The agency reports improving cost per lead. The firm looks at the cases it has signed and does not recognise the picture. Neither party is lying; they are measuring different things, and the agency's metric is the one with the weaker connection to the money.
The other structural difficulty is the intake process. Law firms convert enquiries through a human conversation, and the quality of that conversation determines everything downstream. A firm with excellent marketing and poor intake will produce worse results than a firm with mediocre marketing and a well-run intake team, and no campaign optimisation will close that gap. Agencies who cannot see intake are optimising a funnel whose largest constriction is invisible to them.
The upside is proportionate. An agency that can attribute signed cases to channels, and can show a firm where their intake is losing enquiries, is providing something legal marketing very rarely provides: a number connected to revenue. In a sector with these budgets and these case values, that is worth a great deal and it is defensible in a way that impression share is not.
What goes wrong on legal accounts
These are the specific failures that recur, and most of them are invisible in conventional reporting.
Optimising to the average
Bidding towards cheap enquiries in a sector where one case in fifty carries the year. The metric improves while the revenue does not.
Intake losing the good ones
Enquiries reaching voicemail, being handled by whoever is nearest, or waiting a day for a callback. In legal, a caller who does not reach somebody rings the next firm within minutes.
Windows too short to hold the journey
Family and immigration matters are considered over weeks or months. A 30-day attribution window erases the content that introduced the client and credits the brand search at the end.
Out-of-area and out-of-scope enquiries
Callers in the wrong jurisdiction or needing a practice area the firm does not cover, counted as leads. In some accounts this is a third of everything.
Existing clients in the lead count
Current matters ringing the tracked number for an update. Volume looks healthy, conversion looks poor, and both are artefacts.
Nobody recording outcomes
Without signed-case data, the agency is permanently stuck reporting enquiries while the firm judges them on cases. That gap is where the relationship eventually breaks.
Setting up a law firm account so the numbers mean something
The order matters more here than in most sectors, because the expensive mistakes happen early and compound.
Track every call to source
Legal enquiries are overwhelmingly telephone-first — a person with a legal problem wants to speak to somebody. Dynamic numbers make each call carry its campaign and keyword, which is the foundation everything else stands on.
Classify what the calls actually are
New matter, existing client, out of area, out of scope, wrong number. Until this split exists, the lead count is a mixture of five things and no ratio computed from it is meaningful.
Measure intake before optimising anything
Answer rate, time to answer, out-of-hours volume, callback rate on voicemails. If intake is losing a fifth of enquiries, that is a larger number than anything available in the ad account.
Get the case outcome flowing back
Push enquiries into the firm's case management or CRM so intake works them there, and get signed or not-signed back. This is the step that turns cost per lead into cost per case.
Set a long attribution window
Ninety days as a floor, longer for practice areas where people deliberate. A short window in legal systematically credits brand and defunds everything that built the brand.
Report cases first, enquiries second
Once outcomes exist, lead the report with signed cases by channel. It is the number the firm already uses internally, and speaking their language is most of the battle.
How the practice areas differ
Treating a criminal defence firm and a conveyancing practice as one kind of client is the most common mistake in legal marketing.
| Practice area | Journey | The measurement trap | What to do about it |
|---|---|---|---|
| Personal injury | Fast after the incident, extremely competitive, huge value spread | Cost per lead optimisation in a sector where one case in fifty matters | Case value on signed matters, and accept a high cost per enquiry on the channels that produce serious cases. |
| Family | Long and hesitant — often months of research before contact | Short attribution windows erasing the whole research phase | Ninety days minimum, first-touch credit visible, and content measured on assisted conversions. |
| Criminal defence | Immediate and urgent, frequently out of hours, often a relative calling | Out-of-hours enquiries reaching voicemail | Routing to a duty phone or a legal answering service, then measuring whether it worked. |
| Immigration | Long, research-heavy, multilingual, high repeat and referral | Referral and word-of-mouth work credited to whatever channel was last touched | Honest unattributed reporting, and asking intake to record how people heard of the firm. |
| Employment | Considered, often researched at work on one device and acted on another | Cross-device journeys counted as two unrelated people | Rejoin on identification, and expect a real unattributed share rather than pretending otherwise. |
| Conveyancing / wills | Price-sensitive, comparison-driven, high volume, low value | The one area where cost per lead is genuinely the right metric | Optimise unit cost hard, and separate this reporting from the firm's other practice areas. |
Intake is the largest lever, and it is not in the ad account
Agencies working in legal for any length of time reach the same conclusion: the difference between firms that grow and firms that do not is intake, and it is not close. Two firms with identical marketing budgets, identical rankings and identical enquiry volume can differ by a factor of two in signed cases, entirely on the basis of what happens in the ninety seconds after the phone rings.
The failures are consistent and unglamorous. Calls going to a general reception who takes a message. Enquiries arriving out of hours to voicemail, when a person with a legal emergency is ringing at nine in the evening precisely because it is urgent. Callbacks scheduled for tomorrow, by which point the caller has instructed somebody else. Intake staff who qualify enthusiastically and empathise not at all, on a call where the person is frightened.
None of this is visible to an agency working from platform data, which is why so many legal accounts stall with everybody blaming the wrong thing. With recordings and transcripts it is visible in an afternoon, and the findings tend to be specific enough to act on: eleven calls last month reached voicemail after six, four callers asked about a practice area the firm does cover and were told it did not, the average time to answer is fourteen rings.
Presenting that to a law firm requires some care. Partners are not accustomed to being told their intake is the problem, and a league table of which receptionist performs worst will get recording switched off. The framing that works is volume and pattern rather than individuals: here is when calls are missed, here is what callers are asking for, here is roughly what it is worth. Firms act on that. They do not act on criticism.
The commercial point for the agency is that this work is not marketing and is worth more than marketing. An agency that improves a firm's intake conversion by a fifth has done something no competitor pitching cheaper ad management can replicate, and has become considerably harder to replace.
Compliance, privilege and recording
Legal is a sector where getting this wrong is not merely embarrassing. These are the controls; the advice must come from the firm's own compliance function.
- Recording is per client and off by default. It is never enabled across an account base. A firm that has not agreed to record enquiry calls does not have them recorded.
- An announcement can precede every call. In the firm's own wording, configured per client, played before connection — the standard mechanism in both one-party and two-party consent jurisdictions.
- Retention is set deliberately. Recordings and transcripts expire on a schedule you configure. Holding audio of people describing legal problems indefinitely is a liability the firm will be asked about.
- Capture less on forms. Field-level control means an intake form asking for case details need not have those details stored in the marketing platform at all.
- Deletion is real and per-person. An enquirer asking to be erased can be, individually, and the deletion removes the data rather than hiding it behind a flag.
- Reply carefully in public. Responding to a review can confirm somebody was a client. In legal the safe public reply says nothing about the individual and offers a private channel.
Competing against firms that outspend you
Most law firm accounts are fighting somebody with a bigger budget. In personal injury especially, the top of the market is occupied by national firms and claims companies who can pay more per click than a regional practice can justify, and they will continue to. An agency whose only strategy is bidding is going to lose that fight slowly and expensively.
Better measurement changes the terms of the fight in three specific ways. It lets you find the narrow, unglamorous terms where intent is high and the national spenders are not paying attention — which you can only identify if you can see which keywords produce signed cases rather than which produce cheap enquiries. It lets you shift weight towards organic and the map pack, both of which are disproportionately strong in legal and both of which produce calls rather than forms, meaning they are systematically undercredited by conventional reporting. And it lets you demonstrate that a higher cost per enquiry is correct when those enquiries convert better, which is the argument that stops a nervous client from chasing volume they cannot use.
The intake advantage compounds this. A regional firm cannot outbid a national one, but it can absolutely answer the phone faster and speak to a frightened person better, and those are the two things that convert legal enquiries. An agency that improves the client's conversion of the enquiries they already get is effectively increasing their budget without increasing their spend, which is the only sustainable way to compete against somebody with deeper pockets.
None of this works without the outcome data. The whole argument — bid more here, less there, invest in intake — rests on knowing which enquiries became cases. It is worth being blunt with a firm about that dependency early: without their cooperation on recording outcomes, you are both stuck optimising a proxy, and the proxy is the one that has been misleading legal marketing for twenty years.
Which plan a legal-focused agency needs
Legal agencies tend to run fewer clients at higher value than trades agencies, which changes the calculation. Client count is rarely the binding constraint; call intelligence usually is.
Starter covers a consultant with two or three firms — tracked numbers, recording, form capture and scheduled reports. It has no transcription and no white labelling, which in legal is a real limitation: intake analysis is the highest-value thing you can offer these clients, and it needs transcripts.
Growth is the practical floor for anybody serious about legal. Ten clients, twenty-five numbers, transcription included, white labelling on your own domain, and CRM delivery so enquiries reach the firm's case management system where intake actually works them. That last integration is what eventually produces signed-case data, which is the whole game.
Agency adds AI call intelligence, which in this sector earns its place faster than elsewhere. Automatic classification of new matter against existing client against out-of-area is exactly the split legal accounts need, and doing it by hand across several firms is not sustainable. It also carries the longest data retention, which matters when a practice area has a matter lifecycle measured in years.
One planning note specific to legal: enquiry volumes are lower than trades but each call is worth analysing, so transcription usage is high relative to call count. Budget for analysing most calls rather than a sample, because on a legal account the unanalysed call is the one that was worth six figures.
Bundles that suit a legal book
Longer commitments carry a discount — 15% at three months, 25% at six, 30% annually — which suits legal well, since these are long relationships with predictable volumes.
| If you are | Start on | Because | Add when |
|---|---|---|---|
| A consultant with two or three firms | Starter | Tracked numbers and recording close the attribution gap. Three clients, five numbers, 500 calls a month. | You need transcripts for intake analysis, which is almost immediately. |
| A legal-specialist agency | Growth | Ten clients, transcription for intake findings, white label on your own domain, and CRM delivery into case management so signed-case data comes back. | You need automatic classification across several firms, or pass ten clients. |
| An established legal marketing agency | Agency | Fifty clients, AI classification separating new matters from existing clients and out-of-area callers, and the longest retention for practice areas with multi-year matters. | You start reselling to other agencies — sub-agencies live at this tier. |
The first month on a new legal account
Do three things and hold off on optimising anything until they are done.
Put tracked numbers on the site and let two weeks pass. Then classify what arrived: new matters, existing clients, out of area, out of scope. On most legal accounts the genuine new-matter share is materially lower than the raw enquiry count the firm has been quoted by previous agencies, and establishing the real baseline early prevents a painful correction later.
Second, read fifty calls. Not to grade anybody — to find out where enquiries are being lost. Answer rate, out-of-hours volume, callbacks, and the specific questions callers ask that intake handles badly. This is the meeting that establishes you as something other than an ad manager, and it works best in month one when you have no results of your own to defend.
Third, get the outcome loop started, however crudely. Even a spreadsheet where intake marks signed or not signed against a lead reference is enough to begin computing cost per signed case. It will be incomplete and the coverage should be reported alongside it — "based on the 62% of enquiries with an outcome recorded" — but an incomplete revenue-linked number beats a complete number that is not linked to revenue at all.
Everything after that is ordinary campaign work, made considerably easier by the fact that you now know which enquiries were real, where they came from, and which of them turned into cases.
Common questions
Why is cost per lead a poor metric for law firms?
Because case values vary enormously — one significant matter can be worth more than a hundred routine ones. Optimising towards cheap enquiries reliably improves the metric while reducing the firm's revenue. Cost per signed case is the number that matters, and it requires the firm to record outcomes.
How do I get signed-case data back from a firm?
Deliver enquiries into the case management or CRM system intake already works in, so marking an outcome happens where they already are rather than in your dashboard. Even partial coverage is transformative; report the coverage alongside the figure so the number stays honest.
What attribution window suits legal?
Ninety days as a floor, and longer for family, immigration and employment matters where people deliberate for months. Short windows in legal systematically credit branded search at the end of the journey and defund the content that caused the person to know the firm's name at all.
Can we record enquiry calls given client confidentiality?
That is a question for the firm's compliance function rather than a software vendor. What is provided is the mechanism: recording off by default and enabled per client, a configurable announcement in the firm's own wording before connection, a set retention period, and real per-person deletion.
How do I separate existing clients from new enquiries?
Call classification splits new matters from existing clients, out-of-area callers and out-of-scope enquiries. Until that split exists the lead count is a mixture of five different things, and no conversion rate computed from it means anything.
Is intake really the biggest lever?
In this sector, almost always. Two firms with identical budgets, rankings and enquiry volume routinely differ by a factor of two in signed cases based purely on what happens in the ninety seconds after the phone rings. No campaign optimisation closes a gap that size.
How do I raise intake problems without offending the partners?
Report volume and pattern rather than individuals — when calls are missed, what callers are asking for, roughly what it is worth. A league table of which receptionist scores worst will get recording switched off and usually misdiagnoses a scripting problem as a personal one.
Should practice areas be reported separately?
Yes, always. Conveyancing is high volume and low value; personal injury is the opposite. Averaging them produces a figure that describes nothing that exists and hides the one practice area where the numbers have actually moved.
What plan should a legal-focused agency start on?
Growth is the practical floor, because transcription is what makes intake analysis possible and that is the highest-value thing you can bring a firm. It also carries white labelling and CRM delivery into case management. Agency adds automatic classification, which pays for itself once you run several firms.
How much transcription budget does a legal account need?
Proportionally more than other sectors. Enquiry volumes are lower but every call is worth analysing, so plan for analysing most calls rather than sampling — on a legal account the call you did not analyse is the one that was worth six figures. Usage is metered per client with a ceiling you set.